Abstract
In Brazil and in most of the Latin America countries, the control of the cattle rabies relies upon the indiscriminate culling of the Desmodus rotundus inside a radius of 12 km around an outbreak. The culling is performed by applying warfarin paste in the back of the bats, in the hope that when it returns to its roost after foraging, its conspecifics would ingest the paste during social grooming and die of hemorrhage. This control approach is performed in absentia with the surveillance effort, since if bats die indiscriminately, the effects in the social structure of the colonies are unknown. It is believed that the indiscriminate culling helps the rabies virus spread among bats and consequently, the spillover to livestock. For this reason, any sampling effort in bats and even the identification of cattle rabies outbreaks transmitted by bats is useless to predict future outbreaks or even mitigate current outbreaks. Even though the tendency of the cattle rabies outbreaks is decreasing in Brazil, this should be carefully considered, since rabies surveillance efforts are heterogeneous and decreasing. Even so, the nationwide space-time cattle rabies outbreak shows a wave pattern, increasing the urge of the revision of the current surveillance and control actions. The present work proposes a new approach aiming at decreasing the spillover to livestock by incorporating ecological features of the D. rotundus, most importantly turning the surveillance-control rationale which is currently reactive to livestock rabies cases to a surveillance system reactive to changes in the incidence of vampire bat attacks on the livestock and selective bat population control. For example, estimating the roosts' carrying capacity and contact networks allows the interruption of rabies virus spread by reducing the bat population of target individuals in selected roosts or roost communities. Moreover, the terrain slope dependency of the bat foraging behavior allows to find more efficiently the roost from which bats are attacking a farm or conversely, to find attacked farms if an occupied bat roost is known. These practices could increase the efficiency of the surveillance and control and the cost-effectiveness of the current cattle rabies control program.